Support the Arctic Sea Ice Forum and Blog

Author Topic: Quantifying albedo effect / Rating daily area values  (Read 27354 times)

Tealight

  • Frazil ice
  • Posts: 431
    • View Profile
    • CryosphereComputing
  • Liked: 133
  • Likes Given: 14
Re: Quantifying albedo effect / Rating daily area values
« Reply #50 on: June 21, 2016, 07:35:29 PM »
@tealight - Most informative... Thanks for taking the trouble. One question - does (1 - the overall albedo number for each surface type) map directly to the proportion of solar energy absorbed (given no clouds) or does that need to be summed across the spectrum?
It depends how sophisticated your calculations are. My post is more a comparison between different icetypes all on the same day under the same atmospheric conditions.
As far as I know half of all solar energy is in form of inrared and half as visible/UV light. Then the near infrared is a bit underrepresented and the total albedo should be calculated like: (average(B,G,R)+NIR)/2

I can't quickly find the source now but there are extensive measurements of reflections on sea surfaces and on ice at different angles. The old image from Obuoy3 shows how the water which reflects almost no light straight up to the satellite reflects some light at the incident angle, so does the ice surface, but the snow scatters roughly equally in all directions so that in that case what the satellite sees is representative of the outgoing shortwave radiation, in the case of water and ice it isn't.

Again its more a comparison between ice types to avoid someone believing that only melt ponds are relevant for surface melt.
My surface radiation model considers the solar angle for daily kWh calculations. Without it the daily kWh would be far too high and my sea ice forecast model couldn't work.


Archimid

  • Nilas ice
  • Posts: 2300
    • View Profile
  • Liked: 266
  • Likes Given: 182
Re: Quantifying albedo effect / Rating daily area values
« Reply #51 on: July 06, 2016, 02:47:48 PM »
Fascinating work. Thank you for doing this.

I wonder, do you account for Solar cycles in your calculations? It seems to me that you are taking an averaged value and this being a bottom of solar cycle 24, might make a difference in the predictive value of your model.
I am an energy reservoir seemingly intent on lowering entropy for self preservation.

Tealight

  • Frazil ice
  • Posts: 431
    • View Profile
    • CryosphereComputing
  • Liked: 133
  • Likes Given: 14
Re: Quantifying albedo effect / Rating daily area values
« Reply #52 on: July 06, 2016, 11:24:22 PM »
Fascinating work. Thank you for doing this.

I wonder, do you account for Solar cycles in your calculations? It seems to me that you are taking an averaged value and this being a bottom of solar cycle 24, might make a difference in the predictive value of your model.

Hi Archimid

I don't use solar cycles. According to Wikipedia the total solar irradiance due to solar cycles varies by just 0.1%. What I use is the distance between Earth and Sun. In the Arctic summer solar irradiance is just 96% of the average solar constant.

What I still have to do is switching to NSIDC area like in my "Arctic Sea Ice Forecast Model". Cryosphere Today area is two days behind the actual date which distorts the graphs a little. It also includes fake lake ice generating some random noise.
« Last Edit: July 06, 2016, 11:52:10 PM by Tealight »

bbr2314

  • Nilas ice
  • Posts: 1766
    • View Profile
  • Liked: 160
  • Likes Given: 47
Re: Quantifying albedo effect / Rating daily area values
« Reply #53 on: July 06, 2016, 11:30:51 PM »
Hi Tealight,

Would you be able to update the charts through whatever latest day possible? Very much appreciate the work/effort behind this!!! :)

Archimid

  • Nilas ice
  • Posts: 2300
    • View Profile
  • Liked: 266
  • Likes Given: 182
Re: Quantifying albedo effect / Rating daily area values
« Reply #54 on: July 06, 2016, 11:59:48 PM »
After I made the question I realize you had a development thread. I'm going to reply to you there.
I am an energy reservoir seemingly intent on lowering entropy for self preservation.

Tealight

  • Frazil ice
  • Posts: 431
    • View Profile
    • CryosphereComputing
  • Liked: 133
  • Likes Given: 14
Re: Quantifying albedo effect / Rating daily area values
« Reply #55 on: July 07, 2016, 12:58:48 AM »
OK here are the updated graphs with NSIDC area.

The overall picture hasn't changed, but some years have higher cumulative anomalies than before. Most notably is 2012 with a more significant lead over 2011 and 2007.


Archimid

  • Nilas ice
  • Posts: 2300
    • View Profile
  • Liked: 266
  • Likes Given: 182
Re: Quantifying albedo effect / Rating daily area values
« Reply #56 on: July 07, 2016, 02:44:35 AM »
Since you replied here I'll follow up here.

Looking at the workflow of your algorithms, you multiply(?) the daily irradiance by the area being irradiated. A .1% percent difference might seem small but when applied over such a large area. and over time it is a significant amount of energy. I think that doing so could improve your model more than it seems.
I am an energy reservoir seemingly intent on lowering entropy for self preservation.

Tealight

  • Frazil ice
  • Posts: 431
    • View Profile
    • CryosphereComputing
  • Liked: 133
  • Likes Given: 14
Re: Quantifying albedo effect / Rating daily area values
« Reply #57 on: July 07, 2016, 10:14:10 PM »
Sure 0.1% over the entire arctic is huge amount of energy compared to our electricity usage but not compared to whole icepack.

In April the ice volume is around 23,000 km3
In September it is around 5,000km3

Energy required to melt 1000km3 of ice:
306,866,000,000,000,000 KJ

Additional Energy per day from 0.1% if the whole icepack had meltponds with 0.5 albedo:
       115,400,000,000,000 KJ

So it would take 2659 days of peak summer to melt one additional 1000km3.

In my forecast model I already reduce absorbed energy by 20% because there isn't always clear sky and not all low sea ice concentration directly affects ice melt.

seaicesailor

  • Guest
Re: Quantifying albedo effect / Rating daily area values
« Reply #58 on: July 08, 2016, 01:20:22 AM »
OK here are the updated graphs with NSIDC area.

The overall picture hasn't changed, but some years have higher cumulative anomalies than before. Most notably is 2012 with a more significant lead over 2011 and 2007.

This is amazing work.
May I ask you two questions and make a comment? Do you know the reason behind the large positive anomaly in early 2016? Is it the Atlantic side bare of ice and then the opening of Beaufort? I wanted some confirmatiom to know I understand what is behind such a big anomaly!
Second is if you have an interpretation of what the cummulative plot might mean physically?
The predictions per year look impressive. I assume the more compaction and fram export a given year has, the more your model underpredicts melting. For example, if this year had had more export toward the warmth in the open Atlantic, there would be melting ice unaccounted for in the model. Same happens with compaction (may one come with the other) because there is a lot of heat imported from the Pacific and the continents.
Or maybe the model can somehow grab this; well-devised physically sound models often capture stuff that they did not intend originally (especially and paradoxically when they don't put hundred of effects in the pot).
Forgive these comments, I know they may bother, in my dreams even I could get a tenth of this done. But just for considering. Perhaps I missed something.

Tealight

  • Frazil ice
  • Posts: 431
    • View Profile
    • CryosphereComputing
  • Liked: 133
  • Likes Given: 14
Re: Quantifying albedo effect / Rating daily area values
« Reply #59 on: July 08, 2016, 10:46:06 PM »
I planned a gridded model to do a regional breakdown, but my progress towards it is very slow. For now you have to look at regional area graphs too see where the positive anomaly came from. Beaufort and Barents are probably the main contributors.
https://sites.google.com/site/arcticseaicegraphs/regional

The cummulative plot isn't an exact scientific quantity, but it shows how much more energy went into the arctic compared to other years. Most of 2016 high energy came during spring, when meltponds were absent. So mainly into increasing water temperature in the marginal ice zone or simply prevent refreezing. 2012 on the other hand had high anomalies during June-August. This was caused by melt ponds (high extent/area ratio) so mostly towards ice melt.


Can you clarify your second question a bit more? It is true that my model doesn't explicitly consider fram export or atmospheric forcing. What I have done is fitting the heat from continents to the mid-May to late July additional volume losses. For years which had average Fram export and heat from lower latitudes it works perfectly fine. It doesn't work for all years but the error for the September minimum is always below 10%, which is better than many other forecast models on the Sea Ice Prediction Network.
« Last Edit: July 08, 2016, 11:57:01 PM by Tealight »

seaicesailor

  • Guest
Re: Quantifying albedo effect / Rating daily area values
« Reply #60 on: July 09, 2016, 01:31:01 AM »
Thanks for the explanation!
My question came from the idea that with export, heat external to the system is what melts the exported ice. But it is a bit nonsense, since there is no export external to the regions considered, only ice displacement between regions, e.g. Fram export goes to Greenland sea. Only when breaking up numbers at regional level it would make sense to think on it.


bbr2314

  • Nilas ice
  • Posts: 1766
    • View Profile
  • Liked: 160
  • Likes Given: 47
Re: Quantifying albedo effect / Rating daily area values
« Reply #61 on: July 09, 2016, 01:50:14 AM »
Thanks for the explanation!
My question came from the idea that with export, heat external to the system is what melts the exported ice. But it is a bit nonsense, since there is no export external to the regions considered, only ice displacement between regions, e.g. Fram export goes to Greenland sea. Only when breaking up numbers at regional level it would make sense to think on it.
I think it would make sense when looking at the past month and how ice has shifted within the CAB/peripheral seas. It seems obvious that there has been tremendous movement between the relatively open regions near Siberia and along the Alaskan/Canadian coasts, with the heat pouring into both areas being distributed into the CAB by ice movement.

Perhaps I'm off base but this would also explain why the Siberian Seas are running with cold temp anomalies... if the ice circulating into the area is constantly moving, then the there is much more ice being exposed to the heat, explaining the huge concentration gaps in the CAB.

Archimid

  • Nilas ice
  • Posts: 2300
    • View Profile
  • Liked: 266
  • Likes Given: 182
Re: Quantifying albedo effect / Rating daily area values
« Reply #62 on: July 14, 2016, 01:38:20 AM »
Thank you. Your explanation makes it very clear to me why you do not consider it for your calculations. I wanted to actually see if it could have made a difference.So I attempted to visualize what the maximum posible difference between the solar minimum and maximum would mean for your solar accumulation. I added and subtracted .005 for 184 24 hour days. I didn't even reduce for angles or any other variable and it is already insignificant.

So that's very clear for me now. I still think it might make a difference over yearly and decadal time frames but only a very small one.

I'm going to risk another question.

 I know that the radiation is reduced relative to the tropics because the same amount of sunlight is dispersed over a wider area. However it also means that the light penetrates at a very high angle. The solar energy that dives almost vertically in the tropics, spreads almost horizontally in the arctic.  I wonder if that makes a measurable difference in terms of surface temperatures.

I think figure 4 of this paper illustrates the changes I'm pointing to.

http://onlinelibrary.wiley.com/doi/10.1029/2011GL049421/full
I am an energy reservoir seemingly intent on lowering entropy for self preservation.

Peter Ellis

  • Grease ice
  • Posts: 617
    • View Profile
  • Liked: 35
  • Likes Given: 11
Re: Quantifying albedo effect / Rating daily area values
« Reply #63 on: July 14, 2016, 10:39:06 AM »
Thank you. Your explanation makes it very clear to me why you do not consider it for your calculations. I wanted to actually see if it could have made a difference.So I attempted to visualize what the maximum posible difference between the solar minimum and maximum would mean for your solar accumulation. I added and subtracted .005 for 184 24 hour days.

I'm not clear what the graph adds - but yes, if you reduce the average irradiance by 0.1% to account for a solar minimum, then the cumulative irradiance is also reduced by 0.1%.  It's how maths works.  And 0.1% is not significant in context.

Archimid

  • Nilas ice
  • Posts: 2300
    • View Profile
  • Liked: 266
  • Likes Given: 182
Re: Quantifying albedo effect / Rating daily area values
« Reply #64 on: July 14, 2016, 12:48:03 PM »
Thank you Peter. The purpose behind the graph is to confirm what Tealight and you are saying. Solar cycle variance is insignificant for this analysis. I only add the graph because I wanted to see it with my own eyes and quell the doubt once and for all. Since I did it, I thought I might post it. I know it is extremely simple, but it does convey  important information to me, so maybe it does to someone else?


Edit  For an example of very small amounts becoming significant I like this article:

http://phys.org/news/2016-03-multi-scale-simulations-plasma-turbulence-mystery.html

FTA: "For more than a decade, the general expectation by physicists had been that, because the turbulent "swirls" associated with the ions are so much larger than those associated with electrons, electron-scale swirls would simply be smeared out by the much larger turbulent ion motion. And even if the smaller swirls survived the larger ion-scale turbulence, their tiny size suggested that their effects on heat loss would be negligible.
But the new findings show that this thinking is not always correct. The two scales of turbulence can indeed coexist, the researchers found, and when they do they can interact with each other so strongly that it's impossible to predict the total heat loss accurately unless using a simulation that simultaneously resolves both scales"

I know it is not aplicable for Tealights model, but it might be aplicable to other methods of quantifying albedo effect.
« Last Edit: July 14, 2016, 02:40:30 PM by Archimid »
I am an energy reservoir seemingly intent on lowering entropy for self preservation.

Tealight

  • Frazil ice
  • Posts: 431
    • View Profile
    • CryosphereComputing
  • Liked: 133
  • Likes Given: 14
Re: Quantifying albedo effect / Rating daily area values
« Reply #65 on: July 18, 2016, 12:28:37 AM »
I know that the radiation is reduced relative to the tropics because the same amount of sunlight is dispersed over a wider area. However it also means that the light penetrates at a very high angle. The solar energy that dives almost vertically in the tropics, spreads almost horizontally in the arctic.  I wonder if that makes a measurable difference in terms of surface temperatures.

For water it makes a huge difference because the albedo varies greatly with the angle of incidence. The atmosphere filters out most of the direct radiation as well. If the sun is below 10 degrees over the horizon most radiation is indirect which is about 10% of the total. If earth would be inclined by 30 degrees and not 23.45 we would have much warmer poles and no one would find an ice free arctic extraordinary.

I believe the higher intensity during noon helps to kickstart the formation of melt ponds in the lower latitudes. Even at temperatures slightly below freezing there should be enough energy from sunlight to break down ice crystals.

For my calculations I used aldebo values for smooth water since meltponds or water between ice floes don't have any waves.


Edit: attached is a chart from my model showing insolation over a single day. Considered are the projection effect, atmospheric reductions and water albedo.
« Last Edit: July 18, 2016, 02:36:19 AM by Tealight »

Archimid

  • Nilas ice
  • Posts: 2300
    • View Profile
  • Liked: 266
  • Likes Given: 182
Re: Quantifying albedo effect / Rating daily area values
« Reply #66 on: August 26, 2016, 07:49:37 PM »
https://www.sciencedaily.com/releases/2016/08/160825113235.htm

This article is from a source that I mistrust, but their conclusions make sense to me.

Quote
The effect from Forbush decreases on clouds is too brief to have any impact on long-term temperature changes.

However since clouds are affected by short term changes in galactic cosmic radiation, they may well also be affected by the slower change in Solar activity that happens on scales from tens to hundreds of years, and thus play a role in the radiation budget that determines the global temperature.
I am an energy reservoir seemingly intent on lowering entropy for self preservation.

Tealight

  • Frazil ice
  • Posts: 431
    • View Profile
    • CryosphereComputing
  • Liked: 133
  • Likes Given: 14
Re: Quantifying albedo effect / Rating daily area values
« Reply #67 on: September 03, 2016, 07:20:14 PM »
Here are the updated graphs for "Daily and Cumulative_Energy_Balance_Anomaly".

On the daily graph 2016 stayed close to 2007/11 as one of the most heat absorbing years and on the cumulative graph 2016 stayed in its own league slowly increasing its lead.

Some exact numbers from the Cumulative Energy Balance Anomaly
2012 ended at 264
2007/11 ended at 167
2016 is currently at 374 and will have around 395 at the end of the season.

Thats 50% more than 2012 and 140% more than 2007/11!

I wonder if the cumulative graph could be used to predict the intensity or damaging effects of cyclones at the end of the season. 2012 as second highest year had one great storm and 2016 with a huge lead had a series of powerful storms. I remember that 2013 had a storm as well, but it didn't cause great area/extent losses. Maybe there wasn't enough heat in the ocean. After all it is the second lowest year in the 2007-2016 period.
« Last Edit: September 03, 2016, 07:35:14 PM by Tealight »

Tealight

  • Frazil ice
  • Posts: 431
    • View Profile
    • CryosphereComputing
  • Liked: 133
  • Likes Given: 14
Re: Quantifying albedo effect / Rating daily area values
« Reply #68 on: September 16, 2016, 11:36:06 PM »
Finally after several hours of coding I managed to calculate the albedo anomalies from NSIDC gridded concentration data and present it in an acceptable format. For now I only post the final cumulative anomalies for 2012 and 2016. Later I probably start a new thread where every year is presented in the first comment with external links to gifs or videos for daily anomalies as well.

All anomalies are calculated against the 2007-2016 ice concentration average.
Red indicates lower albedo and above average warming.
Blue indicates higher albedo and below average warming.
7.2 kWh/m2 represents rougly one extra day of peak insulation on open ocean.
220 kWh/m2 represents rougly one extra month of peak insulation on open ocean.
« Last Edit: September 17, 2016, 02:06:21 AM by Tealight »

Neven

  • Administrator
  • First-year ice
  • *****
  • Posts: 7136
    • View Profile
    • Arctic Sea Ice Blog
  • Liked: 697
  • Likes Given: 457
Re: Quantifying albedo effect / Rating daily area values
« Reply #69 on: September 18, 2016, 11:16:13 AM »
You've done some very nice work here, Tealight. My compliments.
Il faut comparer, comparer, comparer, et cultiver notre jardin

Tealight

  • Frazil ice
  • Posts: 431
    • View Profile
    • CryosphereComputing
  • Liked: 133
  • Likes Given: 14
Re: Quantifying albedo effect / Rating daily area values
« Reply #70 on: September 18, 2016, 04:33:16 PM »
You've done some very nice work here, Tealight. My compliments.

Thanks Neven, but you haven't even seen the animations yet. These are even better for analysing the melt season. I still have to assign each pixel its own latitude for individual warming. At the moment every pixel is assigned to 75N, so the data is still preliminary.

Click on the image to start animation

Edit: Animation removed to avoid confusion with new versions
« Last Edit: September 25, 2016, 12:11:20 PM by Tealight »